The CIREP Bellyband Test was designed to simulate the corrosion of refractories at the metal line of aluminum industry's reverberatory furnaces. The test procedure is given below.

The set-up configuration is shown on Fig. 1. It uses a bottom loading type furnace permitting to achieve a temperature drop from about 1150 to 950°C across the air (atmosphere) interface and the molten aluminum alloy. The latter is contained inside the refractory crucible sample which is pre-fired at 1200°C. The crucible sample is cubic with a 7,62 cm edge (a cylinder with the following dimensions may also be used: diameter 7,62 cm, height 7,62 cm). The cavity consists of a hole (3,81 cm diameter x 4,8 cm deep) machined with a diamond tool, thus leaving a 1.9 cm wall minimum thickness. About 90 g of a 5 wt. % Mg alloy is used for each test. This quantity fills two thirds of the cavity. The test is carried out for 4 days (at the soaking temperature).

After the test, the refractory sample is cut longitudinally and two ratings, from 1 (good resistance) to 6 (no resistance), are given according to the extent of metal adherence, penetration and sample friability and/or cracking observed. The first rating is given based on the observation of the portion of the sample that is above the metal line. The second one corresponds to the observation of the sample below the metal line.

* For more information about this section, please consult the following references:

(1) - ALLAIRE, C., "Mechanism of Corundum Growth in Refractories Exposed to Al-Mg Alloys", Aluminum Transactions, Vol. 3, No. 1, October 2000, (2) - ALLAIRE, C., "Refractories for Molten Aluminum Confinement", Journal of the Canadian Ceramic Society, Vol. 69, No. 1, pp. 14-21, February 2000.